- From: Florent Georges <fgeorges@fgeorges.org>
- Date: Thu, 10 Mar 2011 12:10:04 +0100
- To: vojtech.toman@emc.com
- Cc: xproc-dev@w3.org
On 10 March 2011 11:34, ? wrote: > That p:template is a step, whereas p:inline-template is, I > think, meant to be a new type of a binding, like a more dynamic > p:inline. I see, thanks! > But in the end, it was decided to go with a new optional > p:template step. The problem with other approaches was that > they would result in quite a dramatic change to the core > language, and at that time it was simply too late for this. The most problematic one I can see is that it would introduce a change in the way the dependency graph between steps is built. For now, it relies only on step's port bindings, but then it would also rely on some step's port binding's bindings. I am not sure this could be a technical problem, but as a user I feel it would somehow "hide" the dependency (at least in the current state of the language, where it would be an exception with that regard). Regards, -- Florent Georges http://fgeorges.org/
Received on Thursday, 10 March 2011 11:11:00 UTC